Closed mychina75 closed 4 years ago
Hi, thanks for your interest in our work.
Sorry for the confusion as the codes are not well organized yet.
We used the seg-map as in PNG format image, which is 24bit RGB image. The code will automatically map the same sematic between the seg-map of input and the target images.
so, the mask image is a semaintic or panoramic segmentation result? how about the classes defined in the segmentaion? thank you.
Yes, the mask images are the semantic segmented images. The classes defined in image don't matter, as long as values are the same between the source and the target segmentation maps.
I used 24bit RGB image, as in integer value ranged in 0-255. The semantic replacement module will automatically match the same values (class#) between the source and the target segmentation maps.
Hi, thank you for your work. I can not find example to deploy the model. Can you give more infor about the mask image?